The best engineering teams can turn on a dime

The most effective engineering teams I’ve seen share one trait: they can shift priorities suddenly without falling apart.

This sounds simple. It isn’t. Most engineering organisations treat priority changes like emergencies. There’s scrambling, renegotiation, awkward conversations about sunk cost. Engineers feel whiplash. Managers feel like they’re betraying commitments.

But in industries with hot competition, your Q1 objectives might not align with what the market looks like in Q4. A competitor ships something unexpected. A sales conversation reveals a massive opportunity. A customer escalates in ways that change everything. The teams that win are the ones that can respond without ceremony.

Planning still matters

I’m not arguing against planning. Planning is essential. The problem is when planning becomes a ritual that prevents adaptation.

The traditional model looks something like this: spend two weeks at the end of each quarter negotiating priorities across teams, lock everything down, then defend that plan for 90 days regardless of what happens. If something urgent emerges three weeks into the quarter, you’ve got change management overhead, stakeholder escalations, awkward conversations about what gets cut.

According to McKinsey’s 2024 Developer Efficiency Report, engineers spend only 32% of their time writing code¹. The remaining 68% goes to meetings, interruptions and administrative work. Some of that overhead is unavoidable. A lot of it is planning theatre.

The alternative isn’t chaos. It’s treating plans as minimum viable bets. You’re not committing to outcomes for the next 90 days. You’re making the best call you can with current information, then staying ready to adjust when you learn something new.

The CTOs who get this right

Karri Saarinen, CEO of Linear, runs a $1.25B company without OKRs. His team doesn’t set metrics-based goals. They don’t run A/B tests. Decisions are based on taste and opinions, not dashboards.

“We haven’t used OKRs,” Saarinen told Lenny Rachitsky². “For goals, we like to keep it simple and sometimes have more strategic goals, like ‘Be the default tool for startups’ or ‘Get xxx number of companies,’ which we then use as the theme for figuring out the roadmap. I find these types of goals useful to align our team to what we are after without being too specific about how we get there.”

Linear’s teams assemble around projects and disperse when they’re done. No durable cross-functional teams. No product managers for every area. Just small groups that form, ship, and reform. The company has been profitable for over three years with roughly 80 people.

PostHog tells a similar story. They required OKRs in 2022 and walked it back. Engineers were “agonising over finding the right metrics” while feeling like those metrics didn’t reflect their actual progress³. Now teams set goals however they want. Quarterly planning meetings take 60-90 minutes maximum.

“Sometimes we get our goals wrong and that’s okay,” PostHog writes. “Circumstances change, delays happen, engineers need the freedom to adjust.”

When a PostHog team realises they need a different goal, they change it themselves and start working on it immediately. No elaborate change-of-goals process. The priority is shipping valuable products, not perfectly matching the plan you wrote two months ago.

Why engineers resist change

There’s a real human cost when priorities shift. Engineers aren’t being precious when they push back on mid-sprint changes. They’ve invested mental energy in understanding a problem. They’ve built context. Starting fresh means discarding that work and rebuilding context from scratch.

The trick is giving people enough information to understand why the shift matters. When engineers understand the commercial reasoning, when they can see how a pivot connects to real business outcomes, the psychology changes. It stops feeling like wasted effort and starts feeling like responsiveness.

High-functioning teams don’t just tolerate priority shifts. They expect them. The cultural work is making commercial change something people embrace rather than endure.

Dynamic scenario modelling

This is where tooling actually matters.

If changing a plan means regenerating spreadsheets, reconciling headcount across systems and manually recalculating cost impacts, you’ve created structural resistance to adaptation. Every friction point is a reason not to respond to the market.

Being able to quickly assemble a team to build and deploy a pilot, to validate a sales hunch, to test a hypothesis, requires the ability to see the impact of that decision before you commit. What does it cost? What does it do to your other priorities? What’s the trade-off?

This is what we’re building at Flowstate. Move a person from Team A to Team B at 0.6 FTE next month, and cost, FTE and variance recalculate automatically. Clone a budget and compare the impact of a reorg before you commit. The goal isn’t less planning. It’s planning that can evolve without ceremony.

When replanning is cheap, teams can take bigger swings. They can spin up a three-person squad to chase an opportunity, see if it works, then reallocate if it doesn’t. That kind of fluidity isn’t possible when every change requires three weeks of stakeholder alignment.

What the best teams do differently

Linear’s Karri Saarinen puts it simply: “Quality is our first principle. Every other metric and decision flows from that.”⁴

They’ve stayed small and focused. Roughly 80 people at a $1.25B valuation. Profitable since 2021. Only two people have ever left the company. They don’t set aggressive OKRs. They pursue a harder-to-measure metric: craft.

PostHog runs with 26 small teams, each with autonomy to change goals when circumstances demand it. Their philosophy: “It’s better to change a goal to something useful than be stuck working on something useless because you said you would two months ago.”

These companies share a few patterns:

Small teams with real ownership. Not just executing tasks, but deciding what to build and how to build it. Linear’s project teams form around specific work, then disperse when it’s done.
Plans as starting points, not contracts. Goals exist to provide direction, not to lock people into commitments that stop making sense.
Investment in replanning infrastructure. Whether that’s sophisticated feature flags that let you ship incrementally, or workforce planning tools that let you model changes before committing.
Commercial context shared broadly. Engineers understand why priorities shift because leadership explains the business reasoning. Change stops feeling arbitrary.
Tolerance for imperfect plans. As PostHog’s handbook puts it: “All objectives are bad. They have many compromises, are fallible, easy to game, or may be affected by external factors. So use the least bad ones.”

You don’t have to be a startup

The argument isn’t that big companies should run like startups. The argument is that the friction associated with changing direction is often an organisational choice, not an inherent constraint.

Business agility has expanded far beyond tech. Traditional banks launch digital-only services. Manufacturers adopt real-time supply chain adjustments. Government agencies restructure around citizen service delivery. The World Economic Forum argues that resilience and agility alone aren’t sufficient anymore. Organisations need “continuous adaptation” to thrive⁵.

You can have 10,000 employees and still build the cultural muscle to shift quickly. It requires intentional work. It requires investment in tooling. It requires leaders who explain the why behind changes, not just the what.

The market doesn’t care about your quarterly roadmap. It only cares what you ship.